This application addresses broad Challenge Are "(06) Enabling Technologies" and Specific Challenge Topic "06-CA-106 Data integration and visualization methods and tools". Prostate cancer is a highly prevalent disease in older men of the Western world, whose initiation, unregulated growth, invasion and metastasis is characterized by multiple complex events. Gathering genomic, proteomic and metabolomic expression data offers the possibility of deciphering the molecular networks that distinguish progressive from non-progressive forms of the disease;in addition, they provide insight into the biology of aggressive prostate cancer and help in identifying biomarkers that in turn will aid in the selection of patients for treatment. Using prostate cancer as a target application, we propose developing statistical and bioinformatics methodology for the integration and analysis of these three matched data sources (genomic, proteomic and metabolomic expression profiles). These will be complemented by computational and visualization tools to enable robustness in the integrative pipeline. Specifically, a framework for carrying out inference about the identification of enriched pathways from integrated Omics data is developed based on graph theoretic ideas and mixed linear statistical models. Further, the associated estimation of parameters of interest, hypothesis testing and algorithmic issues, are also addressed. Techniques for the identification of genomic/proteomic/metabolomic biomarkers are also discussed. Finally, the scientific findings from the analysis in the form of altered pathways will be validated through testing in associated cell lines and animal models. The results obtained from the proposed studies will provide a suite of computational approaches for integrating diverse Omics data and identify altered pathways in prostate cancer progression. Overall, the results of such integration will enable the use of a combination of molecular markers (genomics, proteomics and metabolomics) as endogenous sensors for cancer progression. Additionally, the methods developed by this proposal for integrative data will advance the field of bioinformatics by building the platform for amalgamation of other types of high throughput Omics data generated using next generation sequencing, comparative genomic hybridization, DNA methylation etc.